JP2016218361A - Speech recognition system, in-vehicle device, and server device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve a response speed to user's speaking while reducing the burden of a communication cost.SOLUTION: If words in user's speaking are not registered with additional dictionary information specifying vehicle devices and operation contents, an in-vehicle device transmits speech information of user's speaking to a server device, and the server device determines whether the speech information acquired from the in-vehicle device intends an operation of the vehicle device or not. If the speech information is determined to intend an operation of the vehicle device as the result of the determination, the server device transmits additional dictionary registration information associating the words in user's speaking and information showing an operation command for the vehicle device with each other, to the in-vehicle device, and the in-vehicle device uses the additional dictionary registration information acquired from the server device to generate additional dictionary information associating the words in user's speaking and the operation command with each other and uses the operation command to output an execution indication for the operation to the corresponding vehicle device.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a voice recognition system, an in-vehicle device, and a server device.

  In Patent Document 1, speech recognition is performed by registering a correspondence list at least associating speech data of speech commands input to the speech input unit with recognition results obtained from the information center based on speech data. If it is determined that it is not possible to perform voice recognition at the information center, based on the utterance data of the voice command input to the voice input unit, A speech recognition system that performs speech recognition on the navigation device side by obtaining a recognition result corresponding to the utterance data using the latest simple dictionary is described.

JP 2010-224301 A

  In the speech recognition system of Patent Document 1, the speech data received from the user is transmitted to the speech recognition server, and the speech recognition result is received, except when the communication between the in-vehicle terminal and the speech recognition server cannot be established. . That is, even when utterance data that matches the user utterance is registered in the simple dictionary, the user's utterance data is transmitted to the voice recognition server except for a predetermined case. However, if the utterance data is transmitted to the voice recognition server each time, there is a problem that the communication time is increased accordingly, the response speed to the user utterance is slowed, and the communication cost is increased.

  Accordingly, an object of the present invention is to provide a speech recognition system that improves the response speed to user utterances while reducing the burden of communication costs.

  In order to solve the above-described problem, a voice recognition system according to the present invention is a voice recognition system including an on-vehicle device and a server device, and the on-vehicle device specifies a vehicle device and an operation content in terms of a user utterance. When not registered in the additional dictionary information, the voice information of the user utterance is transmitted to the server device, and the server device intends the operation of the vehicle device by the voice information acquired from the vehicle-mounted device. And when it is determined that the operation of the vehicle device is intended, an additional dictionary that associates the words of the user utterance with information indicating the operation command of the vehicle device The registration information is transmitted to the on-vehicle device, and the on-vehicle device associates the word of the user utterance with the operation command using the additional dictionary registration information acquired from the server device. Generating the additional dictionary information, by using the operation command, and outputs an instruction to execute the operation with respect to the corresponding said vehicle equipment.

  In addition, the vehicle-mounted device according to the present invention includes a storage unit that stores additional dictionary information in which a word indicating an operation of a vehicle device and an operation command in which an identification number of the vehicle device and an operation content is associated, and a user utterance An input receiving unit that receives input of the voice information, a speech recognition unit that generates speech text information obtained by text-converting the speech information, and searches the additional dictionary information using the speech text information; and a result of the search, A communication unit that transmits voice information of the user utterance to a predetermined server device when a word of the user utterance indicated by the text information is not registered in the additional dictionary information.

The server device according to the present invention further includes an operation command specifying unit that recognizes voice information of a user utterance acquired from the vehicle-mounted device and generates utterance text information obtained by converting the voice information into text. Uses the utterance text information to determine whether or not a word indicated by the user utterance is intended to operate a vehicle device mounted on the vehicle-mounted device.

  According to the voice recognition system of the present invention, it is possible to improve the response speed to user utterance while reducing the burden of communication cost.

  Problems, configurations, effects, and the like other than those described above will be clarified by the following description of embodiments.

It is the figure which showed an example of the hardware constitutions of the onboard equipment which implement | achieves the head unit which concerns on one Embodiment of this invention. It is the figure which showed an example of the hardware constitutions of the server apparatus which concerns on one Embodiment of this invention. It is the figure which showed an example of the functional block of the onboard equipment and server apparatus which concern on one Embodiment of this invention. FIG. 4A is a diagram showing an example of operation command information according to an embodiment of the present invention. FIG. 4B is information indicating an example of additional dictionary information according to an embodiment of the present invention. It is the flowchart which showed an example of the flow of the speech recognition process which concerns on one Embodiment of this invention. It is the flowchart which showed an example of the flow of the operation command specific process which concerns on one Embodiment of this invention. It is the flowchart which showed an example of the flow of the additional dictionary registration process which concerns on one Embodiment of this invention.

  Hereinafter, an embodiment of the present invention will be described.

  The voice recognition (VR) system according to the present embodiment includes a head unit (H / U) and a server device. FIG. 1 is a diagram illustrating an example of a hardware configuration of the vehicle-mounted device 100 that realizes the head unit. The vehicle-mounted device 100 is an electronic device that can be mounted on a vehicle, and includes devices that are distributed in a non-mounted state. Moreover, the onboard equipment 100 which concerns on this embodiment is a navigation apparatus provided with what is called a navigation function, such as a search of a recommended route, and the display of map information and traffic information. However, the vehicle-mounted device 100 is not limited to the navigation device, and any component having a voice input / output function, a voice recognition function, and a communication function with the server device (for example, a vehicle-mounted audio system) may be used. It may be a type of equipment.

  The vehicle-mounted device 100 includes an arithmetic processing device 1, a display 2, a storage device 3, a voice input / output device 4 (having a microphone 41 as a voice input device and a speaker 42 as a voice output device), and an input device. 5, ROM device 6, vehicle speed sensor 7, gyro sensor 8, GPS (Global Positioning System) receiving device 9, FM multiplex broadcast receiving device 10, beacon receiving device 11, communication device 12, CAN I / F (controller area network interface) 13.

  The arithmetic processing device 1 is a central unit that performs various processes of the vehicle-mounted device 100. The arithmetic processing device 1 detects the current location using, for example, various sensors such as the vehicle speed sensor 7 and information output from the GPS receiver 9. The arithmetic processing device 1 reads map information necessary for display from the storage device 3 or the ROM device 6 based on the obtained current location information. Further, the arithmetic processing device 1 develops the read map information in graphics, and outputs a signal to be displayed on the display 2 with a mark indicating the current location superimposed thereon. Further, the arithmetic processing device 1 searches for a recommended route connecting the departure point and the destination instructed by the user using the map information stored in the storage device 3 or the ROM device 6. In addition, the arithmetic processing device 1 outputs a predetermined signal to the speaker 42 and the display 2 to perform route guidance.

  In addition, the arithmetic processing device 1 performs voice recognition processing using voice information of a user utterance input via the microphone 41. In addition, the arithmetic processing device 1 transmits voice information of a user utterance to the server device via the communication device 12 in a predetermined case.

  Further, the arithmetic processing device 1 acquires additional dictionary registration information including text information obtained by text-converting voice information of a user utterance and operation commands for vehicle equipment from the server device via the communication device 12. In addition, the arithmetic processing device 1 generates and updates an additional dictionary using the additional dictionary registration information, and outputs a vehicle device operation execution instruction specified by the operation command.

  The vehicle device is a device that is electrically connected to each other by a CAN (Controller Area Network), such as an air conditioner, an audio, a navigation device, a blinker, and a wiper, or a device that can be operated by a user.

  Such an arithmetic processing device 1 has a configuration in which devices are connected by a bus. Specifically, the arithmetic processing apparatus 1 stores a CPU 21 (Central Processing Unit) that executes various processes such as numerical calculation and control of each device, and map information and arithmetic data read from the storage device 3 or the ROM 23. A RAM 22 (Random Access Memory), a ROM 23 (Read Only Memory) for storing a boot program realized by the CPU 21, a program executed by the CPU 21 (for example, a program realizing a voice recognition (VR) function), and the like, and an arithmetic processing unit 1 includes an I / F 24 (interface) for connecting various kinds of hardware and a bus 25 for connecting them to each other.

  The display 2 is a unit that displays graphics information. The display 2 is composed of, for example, a liquid crystal display or an organic EL display.

  The storage device 3 includes at least a readable / writable storage medium such as an HDD (Hard Disk Drive) or a nonvolatile memory card. In the storage device 3, for example, various information (for example, map information) used by the arithmetic processing device 1 is stored.

  The voice input / output device 4 includes a microphone 41 as a voice input device and a speaker 42 as a voice output device. The microphone 41 acquires sound outside the in-vehicle device 100 such as a voice (user utterance) uttered by a driver or a passenger. Further, the speaker 42 outputs the guidance to the driver and the like generated by the arithmetic processing device 1 as voice.

  The input device 5 is a device that receives an instruction input from a user. The input device 5 includes a touch panel 51, a dial switch 52, scroll keys that are other hard switches, and the like (not shown). The input device 5 outputs information corresponding to the operation of each key or each switch to other devices such as the arithmetic processing device 1.

  The ROM device 6 is composed of at least a readable storage medium such as a ROM such as a CD-ROM or a DVD-ROM, or an IC (Integrated Circuit) card. In this storage medium, for example, moving image data, audio data, and the like are stored.

  The vehicle speed sensor 7 is a sensor that outputs a value used to calculate the vehicle speed. The gyro sensor 8 is configured by an optical fiber gyro, a vibration gyro, or the like, and is a sensor that detects an angular velocity due to rotation of a moving body. The GPS receiver 9 receives a signal from a GPS satellite and measures the distance between the mobile body and the GPS satellite and the rate of change of the distance with respect to three or more satellites to thereby determine the current location, travel speed, and travel of the mobile body. It measures the direction. Each of these devices is used in an arithmetic processing unit to detect the current location of a vehicle on which the vehicle-mounted device 100 is mounted.

  The FM multiplex broadcast receiver 10 receives an FM multiplex broadcast signal transmitted from an FM broadcast station. FM multiplex broadcasting includes VICS (Vehicle Information Communication System) information, as well as current traffic information, regulation information, SA / PA (service area / parking area) information, parking information, weather information, and FM multiplex general information. Contains text information provided by radio stations.

  The beacon receiving device 11 receives rough current traffic information such as VICS information, regulation information, SA / PA (service area / parking area) information, parking lot information, weather information, emergency alerts, and the like. The beacon receiving device 11 includes, for example, an optical beacon that communicates by light and a radio beacon that communicates by radio waves.

  The communication device 12 performs information communication with an external device (a server device in this example). Specifically, the communication device 12 transmits the voice information of the user utterance to a predetermined server device. Further, the communication device 12 receives additional dictionary registration information from the server device.

  The CAN I / F 13 is an interface for inputting / outputting information to / from an in-vehicle network (CAN) that performs multiplex communication with various vehicle devices installed in the vehicle. The CAN I / F 13 may be realized by the I / F 24 described above.

  The hardware configuration of the vehicle-mounted device 100 has been described above.

  Next, the hardware configuration of the server device 200 will be described. FIG. 2 is a diagram illustrating an example of a hardware configuration of the server device 200. The server apparatus 200 is an information processing apparatus such as a workstation or a PC (personal computer).

  As illustrated, the server device 200 includes an arithmetic device 201, an external storage device 202, a transmission / reception device 203, and a bus 204 that connects the respective devices to each other.

  The arithmetic device 201 is realized by the CPU 211 that executes various processes such as numerical calculation and control of each device, the RAM 212 that stores map information and arithmetic data read from the external storage device 202 or the ROM 213 described later, and the CPU 211. It has a ROM 213 that stores a boot program, a program executed by the CPU 211 (for example, a program that realizes a voice recognition (VR) function), and a bus 204 that interconnects the devices.

  The external storage device 202 is a nonvolatile storage device such as a hard disk device or a flash memory.

  The transmission / reception device 203 is, for example, a device such as a communication module that performs information communication with an external device (in this example, the vehicle-mounted device 100).

  The hardware configuration of the server device 200 has been described above.

  Next, functional blocks showing functional configurations of the vehicle-mounted device 100 and the server device 200 will be described. FIG. 3 is a diagram illustrating an example of functional blocks of the vehicle-mounted device 100 and the server device 200. The vehicle-mounted device 100 includes an input receiving unit 301, an output processing unit 302, a voice recognition unit 303, an additional dictionary registration unit 304, a storage unit 305, and a communication unit 306.

  The input receiving unit 301 is a functional unit that receives an instruction or information input from the user via the input device 5 provided in the vehicle-mounted device 100. Specifically, the input receiving unit 301 receives a voice recognition processing execution instruction from the user via the input device 5 such as a hard switch. More specifically, when the input reception unit 301 detects that a predetermined hard switch has been pressed by the user, the input reception unit 301 notifies the voice recognition unit 303 of this and starts a PTT (Push To Talk) function.

  The output processing unit 302 is a functional unit that generates screen information to be displayed on the display 2 included in the vehicle-mounted device 100. Specifically, the output processing unit 302 generates screen information indicating the result of voice recognition and outputs the screen information to the display 2. For example, when the user utterance is a display request for a specific place such as “restaurant” or “bank”, the output processing unit 302 acquires coordinate information from the map information, and overlays an icon indicating the place on the map. Screen information to be displayed is generated and output to the display 2.

  The speech recognition unit 303 is a functional unit that performs speech recognition of input speech. Specifically, when the voice recognition unit 303 acquires voice information of a user utterance input via the microphone 41, the voice recognition unit 303 performs acoustic analysis using an acoustic model, and performs voice recognition processing of the input voice using a VR dictionary. Run. Also, the voice recognition unit 303 converts voice information of the user utterance into text information as a result of the voice recognition process. Note that the voice recognition processing method is not particularly limited, and a known voice recognition technique may be used.

  The additional dictionary registration unit 304 is a functional unit that generates and updates additional dictionary information 315. Specifically, the additional dictionary registration unit 304 uses the additional dictionary registration information acquired from the server device 200 via the communication unit 306 to generate and update additional dictionary information 315.

  The storage unit 305 is a functional unit that stores various information. Specifically, the storage unit 305 includes map information 311, a VR dictionary 312, operation command information 313, standby vocabulary information 314, and additional dictionary information 315.

  The map information 311 is road configuration information storing mesh area information including link information related to roads on the map.

  The VR dictionary 312 is dictionary information in which phonemes and words are registered in association with each other, and is used when the speech recognition unit 303 performs speech recognition processing using speech information of user utterances.

  The operation command information 313 is information related to the operation of the vehicle device. FIG. 4A is a diagram illustrating an example of the operation command information 313. The operation command information 313 is composed of a command group 1 and a command group 2, information specifying a vehicle device is registered in the command group 1, and the operation content of the vehicle device specified in the command group 1 is registered in the command group 2. Information to be identified is registered.

  Specifically, in the command group 1, an identification number that is identification information of a vehicle device and a vehicle device name are registered in association with each other. For example, information such as “1. air conditioner”, “2. audio”, and “3. navigation” is registered in the command group 1, and “1.” to “3.” are vehicle device identification numbers. . “Air conditioner”, “audio” and “navigation” are vehicle device names.

  In the command group 2, for each vehicle device specified in the command group 1, an operation identification number and an operation content are registered in association with each other. For example, the command group 2 associated with “1. air conditioner” in the command group 1 includes “1. ON”, “2. OFF”, “3. Increase temperature”, “4. Lower temperature”, Information such as “5. High air flow” and “6. Low air flow” is registered. Here, “1.” to “6.” are operation identification numbers. “ON”, “increase temperature”, and the like are operation details.

  An operation command for each vehicle device is specified by a combination of identification numbers of the command group 1 and the command group 2. For example, the operation command in the case of “decrease the temperature of the air conditioner” is specified by a combination of identification numbers (command group 1 = 1, command group 2 = 4).

  Returning to FIG. The standby vocabulary information 314 is information in which a predetermined vocabulary is registered. For example, predetermined words such as “Destination”, “Home”, “Brightness”, “Map (MAP)”, “Route”, “Restaurant”, and “Bank” are registered in the standby vocabulary information 314. And is used as a search target of the utterance text information generated by the voice recognition unit 303.

  The additional dictionary information 315 is information in which a user's utterance voice, which is text information, and an operation command of the vehicle device are associated with each other. FIG. 4B is information indicating an example of the additional dictionary information 315. Specifically, the additional dictionary information 315 has a record in which the speech voice field 321, the command group 1 field 322, and the command group 2 field 323 are associated with each other.

  Information registered in the utterance voice column 321 is information indicating the contents of the user utterance text-converted by the voice recognition unit 303. In the utterance voice column 321, for example, information indicating the contents of the user utterance such as “cold”, “hot”, “lower the air conditioner” is registered. The information registered in the command group 1 column 322 is an identification number that identifies the vehicle device, and corresponds to the identification number of the command group 1 in the operation command information 313. The information registered in the command group 2 column 323 is an operation identification number that identifies the operation content of the vehicle device, and corresponds to the operation identification number of the command group 2 in the operation command information 313. Such additional dictionary information 315 is generated by the additional dictionary registration unit 304 using the additional dictionary registration information acquired from the server device 200.

  Returning to FIG. The communication unit 306 is a functional unit that exchanges information with an external device (the server device 200 in this example). Specifically, the communication unit 306 transmits voice information of user utterances to a predetermined server device 200 via a predetermined network N such as the Internet. Further, the communication unit 306 receives additional dictionary registration information including text information of user utterances and operation commands of vehicle equipment from the server device 200 via the network N.

  The server device 200 includes an operation command specifying unit 401, an output unit 402, an additional dictionary registration information generation unit 403, an information storage unit 404, and a transmission / reception unit 405.

  The operation command specifying unit 401 is a functional unit that specifies an operation command of a vehicle device using voice information of a user utterance. Specifically, when the operation command specifying unit 401 acquires the voice information of the user utterance from the vehicle-mounted device 100 via the transmission / reception unit 405, the operation command specifying unit 401 performs a voice recognition process and converts the voice information of the user utterance into text information. In addition, the operation command specifying unit 401 uses the text information to determine whether or not the user utterance is intended to operate the vehicle device. Further, the operation command specifying unit 401 uses the operation command information 414 to specify the operation command of the vehicle device indicated by the user utterance.

  The output unit 402 is a functional unit that specifies information obtained by user utterance and transmits the information to the vehicle-mounted device 100 via the transmission / reception unit 405. Specifically, the output unit 402 specifies a point coordinate using the map information 411 when the user utterance is not a vehicle device operation command but a display request for a specific point, and the transmission / reception unit 405 is used. To the vehicle-mounted device 100.

  The additional dictionary registration information generation unit 403 is a functional unit that generates information registered in the additional dictionary information of the in-vehicle device 100. Specifically, the additional dictionary registration information generation unit 403 generates additional dictionary registration information including text information of user utterances and operation commands for vehicle equipment, and transmits the additional dictionary registration information to the in-vehicle device 100 via the transmission / reception unit 405.

  The information storage unit 404 is a functional unit that stores various information. Specifically, the information storage unit 404 stores map information 411, a VR dictionary 412, a context dictionary 413, and operation command information 414. Note that the map information 411, the VR dictionary 412 and the operation command information 414 are the same as those stored in the storage unit 305 of the vehicle-mounted device 100, and thus the description thereof is omitted.

  The context dictionary 413 is dictionary information in which associations between words corresponding to the context are registered, and is used when the operation command specifying unit 401 performs speech recognition processing using speech information of user utterances.

  The functional blocks of the vehicle-mounted device 100 and the server device 200 have been described above. Note that the input receiving unit 301, the output processing unit 302, the voice recognition unit 303, and the additional dictionary registration unit 304 of the vehicle-mounted device 100, the operation command specifying unit 401, the output unit 402, and the additional dictionary registration information generation unit 403 of the server device 200, Is realized by a program for causing the CPU 21 and the CPU 211 to perform processing. Each of these programs is stored in the ROM 23 or the storage device 3 of the vehicle-mounted device 100 and the ROM 213 or the external storage device 202 of the server device 200. The programs are loaded onto the RAM 22 and the RAM 212 and executed by the CPU 21 and the CPU 211. Note that the VR dictionary 312 of the vehicle-mounted device 100 is not limited to being stored in the storage device 3, and may be stored in the ROM 23. In addition, the VR dictionary 412 and the context dictionary 413 of the server device 200 are not limited to being stored in the external storage device 202, and may be stored in the ROM 213.

  Each functional block is classified according to the main processing contents in order to facilitate understanding of the functions of the vehicle-mounted device 100 and the server device 200 realized in the present embodiment. Therefore, the present invention is not limited by the classification method of each function or its name. Moreover, each structure of the onboard equipment 100 and the server apparatus 200 can also be classified into many more components according to the processing content. Moreover, it can also classify | categorize so that one component may perform more processes.

  In addition, all or part of each functional unit may be constructed by hardware (an integrated circuit such as an ASIC) mounted on a computer. Further, the processing of each functional unit may be executed by one hardware, or may be executed by a plurality of hardware.

  Note that the storage unit 305 of the vehicle-mounted device 100 is realized by the storage device 3. Further, the communication unit 306 of the vehicle-mounted device 100 is realized by the communication device 12. Further, the storage unit 305 of the server device 200 is realized by the external storage device 202. The transmission / reception unit 405 of the server device 200 is realized by the transmission / reception device 203.

[Description of operation]
Next, the speech recognition process performed with the onboard equipment 100 is demonstrated. FIG. 5 is a flowchart showing an example of the flow of voice recognition processing. When the input receiving unit 301 receives a voice recognition processing start instruction from the user via the input device 5, the input receiving unit 301 starts such processing.

  When the voice recognition process is started, the input receiving unit 301 determines whether or not a voice input of a user utterance has been received via the microphone 41 (step S001). In addition, when the input reception unit 301 receives a voice input of a user utterance (Yes in step S001), the input reception unit 301 transfers the input voice information to the voice recognition unit 303.

  The speech recognition unit 303 performs speech recognition processing using a predetermined acoustic model or VR dictionary, and generates speech text information obtained by text-converting the speech information of the user utterance acquired through the input reception unit 301 (step S002). .

  Next, the voice recognition unit 303 determines whether or not the user utterance words converted into the utterance text information are registered in the standby vocabulary information 314 (step S003). Specifically, the voice recognition unit 303 searches the standby vocabulary information 314 using the utterance text information. When the words of user utterances indicated by the utterance text information are registered in the standby vocabulary information 314 (Yes in step S003), the output processing unit 302 executes a predetermined process corresponding to the searched standby vocabulary ( Step S004). For example, when the word of the user utterance indicated by the utterance text information is “bank”, the output processing unit 302 acquires the point coordinates of the bank around the current location using the map information, and displays a mark indicating the point on the map. The screen information superimposed on is generated. The output processing unit 302 outputs the generated screen information to a display.

  On the other hand, when the word of the user utterance indicated by the utterance text information is not registered in the standby vocabulary information 314 (No in step S003), the speech recognition unit 303 determines whether the word is registered in the additional dictionary information 315. Is determined (step S005). Specifically, the voice recognition unit 303 searches the additional dictionary information 315 using the utterance text information.

  When the user utterance word indicated by the utterance text information is registered in the additional dictionary information 315 (Yes in step S005), the voice recognition unit 303 outputs an operation instruction for the vehicle device (step S006). Specifically, the voice recognition unit 303 specifies a record of the additional dictionary information 315 associated with the utterance voice that matches the utterance text information. Further, the voice recognition unit 303 refers to the operation command information 313 using the identification numbers stored in the command group 1 column 322 and the command group 2 column 323 of the specified record, and determines the target vehicle device and the operation content. Identify. Further, the voice recognition unit 303 outputs an execution instruction for the specified operation content to the specified vehicle device via the CAN I / F 13.

  On the other hand, when the word of the user utterance indicated by the utterance text information is not registered in the additional dictionary information 315 (No in step S005), the voice recognition unit 303 uses the voice information acquired from the input reception unit 301 as the communication unit 306. To the server apparatus 200, and the process of this flow is terminated.

  In the above, the speech recognition process performed with the onboard equipment 100 was demonstrated.

  Next, the operation command specifying process of the server device 200 that acquires the voice information of the user utterance from the vehicle-mounted device 100 will be described. FIG. 6 is a flowchart showing an example of the flow of the operation command specifying process. Such processing is started when the operation command specifying unit 401 of the server device 200 acquires the voice information of the user utterance from the in-vehicle device 100 via the transmission / reception unit 405. That is, the command specifying process 401 is performed when the words of user utterances are not registered in the standby vocabulary information 314 and the additional dictionary information 315 of the in-vehicle device 100.

  When the voice information of the user utterance is acquired, the operation command specifying unit 401 of the server device 200 performs voice recognition processing using a predetermined acoustic model, the VR dictionary 412 and the context dictionary 413, and converts the voice information of the user utterance into text. Utterance text information is generated (step S011).

  Next, the operation command specifying unit 401 uses the utterance text information to determine whether or not the voice information of the user utterance indicates an operation of the vehicle device (step S012). Specifically, the operation command specifying unit 401 uses the utterance text information to determine whether the content of the user utterance is intended for the operation of the vehicle device or otherwise. When it is determined that the operation of the vehicle device is not intended (No in step S012), the operation command specifying unit 401 executes a predetermined process corresponding to the voice information (step S013). For example, when the utterance text information is “three banks in this area”, the operation command specifying unit 401 acquires the point coordinates of the three banks closest to the current location from the map information 411, and transmits via the transmission / reception unit 405. It transmits to the onboard equipment 100.

  On the other hand, when it is determined that the content of the user utterance is intended to operate the vehicle device (Yes in step S012), the operation command specifying unit 401 specifies the operation command of the vehicle device using the operation command information 414. (Step S014). For example, when the utterance text information is “hot”, the operation command specifying unit 401 specifies that the vehicle device to be operated is an “air conditioner” that adjusts the room temperature, and assigns the identification number “1” from the command group 1. Identify. Further, the operation command specifying unit 401 specifies that the operation content is “decrease temperature” from the utterance text information, and specifies the operation identification number “4” from the command group 2. In this way, the operation command specifying unit 401 specifies the operation commands “1” and “4” corresponding to the user utterance “hot”.

  Next, the additional dictionary registration information generation unit 403 generates additional dictionary registration information (step S015) and transmits it to the vehicle-mounted device 100 via the transmission / reception unit 405 (step S016). Specifically, the additional dictionary registration information generation unit 405 generates additional dictionary registration information including utterance text information generated based on the voice information of the user utterance and the operation command of the vehicle device specified using the utterance text information. Generated and transmitted to the vehicle-mounted device 100 via the transmission / reception unit 405.

  The vehicle device operation command specifying process executed by the server device 200 has been described above.

  Next, an additional dictionary registration process of the vehicle-mounted device 100 that acquires additional dictionary registration information from the server device 200 will be described. FIG. 7 is a flowchart showing an example of the flow of additional dictionary registration processing. Such processing is started when the additional dictionary registration unit 304 of the vehicle-mounted device 100 acquires additional dictionary registration information from the server device 200 via the communication unit 306.

  The additional dictionary registration unit 304 of the in-vehicle device 100 generates and updates the additional dictionary information 315 using the acquired additional dictionary registration information (step S021). Specifically, the additional dictionary registration unit 304 extracts utterance text information from the additional dictionary registration information and stores it in the utterance voice column 321 of the additional dictionary information 315. Further, the additional dictionary registration unit 304 extracts operation commands from the additional dictionary registration information and stores them in the command group 1 column 322 and the command group 2 column 323, respectively. For example, in this example, the additional dictionary registration unit 304 stores the utterance text information “hot” in the utterance voice column 321 of the additional dictionary information 315, and the operation commands “1” and “4” respectively in the command group 1 column. 322 and command group 2 column 323.

  Next, the voice recognition unit 303 outputs an execution instruction for the operation of the vehicle device specified by the command group 1 and the command group 2 to the corresponding vehicle device (step S022). Specifically, the voice recognition unit 303 identifies the record of the additional dictionary information 315 generated and updated by the additional dictionary registration unit 304. In addition, the voice recognition unit 303 identifies the identification numbers stored in the command group 1 column 322 and the command group 2 column 323 of the identified record, and uses the identification number to identify the target vehicle device from the operation command information 313. And identify the operation. In addition, the voice recognition unit 303 outputs an execution instruction for the specified operation content to the specified vehicle device. In this example, the voice recognition unit 303 outputs an instruction to lower the temperature to an air conditioner that is a target vehicle device. Further, when the voice recognition unit 303 outputs the execution instruction, the process of this flow ends.

  The additional dictionary registration process has been described above.

  According to such a voice recognition system, even when there is a user utterance instructing the operation of the vehicle device, if the operation command for the vehicle device is registered in the additional dictionary information, the voice information is sent to the server device 200. Even if it does not transmit and acquire the operation command, the vehicle apparatus can specify the operation command on the vehicle-mounted device 100 side. Thereby, in the speech recognition system according to the present invention, the opportunity to transmit speech information to the server device 200 can be reduced, so that the communication cost burden can be reduced. Since the operation command can be specified on the vehicle-mounted device 100 side after the next time, the response time from the server device 200 can be omitted, and the response speed to the user utterance can be improved.

  The voice recognition system also recognizes voice information of user utterances, and generates additional dictionary information in which operation commands for vehicle equipment are associated with text information of user utterances. Therefore, one common operation command can be associated with each of a plurality of words having different expressions. By using such additional dictionary information, the vehicle-mounted device 100 can identify a single operation command by absorbing the habit of the user's wording. That is, when instructing the operation of the vehicle device, the user does not need to remember a specific word as the operation command.

  Usually, the words registered in the standby vocabulary information are also updated with the update of the map information. However, in the present invention, the additional dictionary information in which the operation commands of the vehicle equipment are registered is obtained from the standby vocabulary information. Because they are stored independently, they are not affected by such updates.

  The present invention is not limited to the above-described embodiment, and various modifications can be made. For example, when updating the additional dictionary information 315, the storage unit 305 of the vehicle-mounted device 100 determines whether or not the information registered in the additional dictionary information 315 is greater than or equal to a predetermined number (for example, 100 or more). When it is determined that the number is greater than or equal to the number, the vocabulary that is not frequently used may be deleted from the standby vocabulary information 314 to increase the usage area of the additional dictionary information 315.

  Further, the storage unit 305 may delete the utterance voices and operation commands registered in the additional dictionary information 315 in order from the least frequently used information when the predetermined number or more (for example, 100 or more).

  According to the vehicle-mounted device of such a speech recognition system, the usage area of the additional dictionary information 315 can be increased as necessary.

  Further, the present invention is not limited to the above-described embodiments and modifications, and includes various embodiments and modifications in addition to these. For example, the above embodiment has been described in detail for easy understanding of the present invention, and is not necessarily limited to the one having all the configurations described. In addition, a part of the configuration of an embodiment can be replaced with the configuration of another embodiment or a modification, and the configuration of another embodiment can be added to the configuration of a certain embodiment. In addition, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

  In addition, each or all of the above-described configurations, functions, processing units, processing means, and the like may be realized by a program in which the processor realizes each function. Information such as programs, tables, and files for realizing each function is stored in a memory, a storage device such as a hard disk or SSD (Solid State Drive), an IC card, an SD (Secure Digital) memory card, or a DVD (Digital Versatile Disk). Can be placed on a recording medium. Note that the control lines and information lines are those that are considered necessary for the explanation, and not all control lines and information lines on the product are necessarily shown.

DESCRIPTION OF SYMBOLS 100 ... Onboard equipment, 1 ... Arithmetic processing device, 2 ... Display, 3 ... Memory | storage device, 4 ... Voice input / output device, 41 ... Microphone, 42 ... Speaker,
5 ... input device, 51 ... touch panel, 52 ... dial switch,
6 ... ROM device, 7 ... vehicle speed sensor, 8 ... gyro sensor,
9 ... GPS receiver, 10 ... FM multiplex broadcast receiver, 11 ... beacon receiver,
12 ... Communication device, 13 ... CAN I / F,
200: server device, 201: arithmetic device, 211: CPU,
212 ... RAM, 213 ... ROM, 202 ... external storage device,
203 ... Transceiver, 204 ... Bus

Claims (10)

A speech recognition system including an on-vehicle device and a server device,
The in-vehicle device is
When the words of the user utterance are not registered in the additional dictionary information that identifies the vehicle device and the operation content, the voice information of the user utterance is transmitted to the server device,
The server device
When it is determined whether the audio information acquired from the vehicle-mounted device is intended for operation of the vehicle device, and, as a result of the determination, when it is determined that the operation of the vehicle device is intended, the user Sending additional dictionary registration information that associates the words of the utterance and information indicating the operation command of the vehicle device to the vehicle-mounted device,
The in-vehicle device is
Using the additional dictionary registration information acquired from the server device, to generate additional dictionary information that associates the words of the user utterance and the operation command,
An operation execution instruction is output to the corresponding vehicle device using the operation command. The speech recognition system according to claim 1,
The in-vehicle device is
A storage unit that stores additional dictionary information in which words indicating the operation of the vehicle device and an operation command combining the identification information of the vehicle device and the operation content are associated;
An input receiving unit that receives input of voice information of a user utterance;
Generating speech text information obtained by text-converting the speech information, and searching the additional dictionary information using the speech text information;
As a result of the search, when a word of a user utterance indicated by the text information is not registered in the additional dictionary information, a communication unit that transmits voice information of the user utterance to the server device,
The server device
It includes an operation command specifying unit that recognizes voice information of a user utterance acquired from the in-vehicle device and generates utterance text information obtained by converting the voice information into text,
The operation command specifying unit includes:
A speech recognition system for determining whether or not a word indicated by the user utterance is intended to operate a vehicle device mounted on the vehicle-mounted device, using the utterance text information. The speech recognition system according to claim 2,
The server device
An information storage unit storing operation command information in which the vehicle device and operation content are associated with identification information;
An additional dictionary registration information generating unit for generating additional dictionary registration information including information related to the operation command of the vehicle device,
The operation command specifying unit includes:
When it is determined that the word indicated by the user utterance is intended to operate the vehicle device, the operation command of the vehicle device is identified using the operation command information,
The additional dictionary registration information generation unit
Sending additional dictionary registration information in which the operation command is associated with the utterance text information to the in-vehicle device,
The in-vehicle device is
Voice recognition, further comprising: an additional dictionary generation unit that generates and updates the additional dictionary information using the additional dictionary registration information when the additional dictionary registration information is acquired from the server device via the communication unit system. The speech recognition system according to claim 3,
The voice recognition unit
Using the utterance text information, the additional dictionary information, and the operation command information, specify the target vehicle device and the operation content,
A voice recognition system that outputs an instruction to execute the operation content to the identified vehicle device. A storage unit that stores additional dictionary information in which words indicating the operation of the vehicle device and an operation command combining the identification information of the vehicle device and the operation content are associated;
An input receiving unit that receives input of voice information of a user utterance;
Generating speech text information obtained by text-converting the speech information, and searching the additional dictionary information using the speech text information;
A communication unit that transmits voice information of the user utterance to a predetermined server device when a word of the user utterance indicated by the text information is not registered in the additional dictionary information as a result of the search. Onboard equipment. The vehicle-mounted device according to claim 5,
The storage unit
It further has operation command information that associates vehicle equipment and operation content with identification information,
The voice recognition unit
When words of user utterances indicated by the utterance text information are registered in the additional dictionary information, using the additional dictionary information and the operation command information, specify a target vehicle device and operation content,
An in-vehicle device that outputs an instruction to execute the operation content to the identified vehicle device. The vehicle-mounted device according to claim 6,
The vehicle-mounted device further comprising: an additional dictionary generation unit that generates and updates the additional dictionary information using the additional dictionary registration information when additional dictionary registration information is acquired from the server device via the communication unit . The vehicle-mounted device according to claim 7,
The storage unit
When updating the additional dictionary information, it is determined whether or not the information registered in the additional dictionary information is greater than or equal to a predetermined number. If it is determined that the information is greater than or equal to the predetermined number, the standby where the predetermined vocabulary is registered is determined. A vehicle-mounted device that deletes vocabulary that is less frequently used from vocabulary information. It includes an operation command specifying unit that recognizes voice information of a user utterance acquired from the in-vehicle device and generates utterance text information obtained by converting the voice information into text,
The operation command specifying unit includes:
The server apparatus characterized by using the said utterance text information, and determining whether the word which the said user utterance intends operation of the vehicle equipment mounted in the said onboard equipment. The server device according to claim 9,
An information storage unit storing operation command information in which the vehicle device and operation content are associated with identification information;
An additional dictionary registration information generating unit for generating additional dictionary registration information including information related to the operation command of the vehicle device,
The operation command specifying unit includes:
When it is determined that the word indicated by the user utterance is intended to operate the vehicle device, the operation command of the vehicle device is identified using the operation command information,
The additional dictionary registration information generation unit
A server device that generates additional dictionary registration information in which the operation command is associated with the utterance text information.

Images